Languages: English | Français
A poisoning attack that causes a model to perform an adversary-selected behaviour in response to inputs that follow a particular backdoor pattern.
Source: NIST AI 100-2e2025 | Category:
